Chapter 73 Types of Communication

At this point in the Course Set, you’ve mastered Google Products and version control with GitHub. You’ve become comfortable working with data in R, understand the importance of tidy data, and know how to visualize your data to better understand it. These are all incredibly important skills for a data scientist; however, having these skills is simply not enough. Another crucial skill set for a data scientist to master is that of communication. The most interesting results in the world are only the most interesting results in the world if the world knows about them.

To ensure that your interesting results move effectively from your brain and computer to the brains of the rest of the world, we’re dedicating an entire course to data science communication. In this course we’ll discuss the various ways in which data scientists regularly communicate and provide details on how to master each of these types of communication. We’ll start in this lesson by providing an overview of the many types of communication. The rest of the lessons will then describe in detail the considerations to make when executing each of these types of communication.

73.0.1 Opinionated Communication

Before we dive into the various types of data science communication, let’s take just a second and talk about opinionated software development. When people talk about software being opinionated, it means that the piece of software forces you to do things their way. For example, if you’re familiar with the company Apple, they develop opinionated products. For example, on an iPhone, the general appearance from one iPhone to the next is very similar and there is often only one way to accomplish a task (for example, to open an app, you click on the app on the touch screen). This simplifies things for the user, since there’s generally only one way to accomplish a task. It also allows one iPhone user to quickly figure out how to work someone else’s iPhone. Alternatively, un-opinionated software is a lot more flexible. Un-opinionated software allows developers to accomplish goals in many different ways, allowing different styles and approaches to be implemented.

Let’s use an arrow as our example. Say I was teaching you to draw an arrow. If I only allowed you to draw one type of arrow, that would be an opinionated approach. If, however, you were able to draw lots of different arrows, that would be an un-opinionated approach. While the second approach is more flexible and provides more options, simply knowing how to draw one arrow is enough to get you started!

For the purposes of the lessons in this course, we’re going to be very opinionated. By using this approach, we can provide you with the necessary information for effective data science communication without providing tons of caveats or confusing alternative approaches. We’re going to provide you with tips and details that have worked for us over the years as data scientists. However, this does not mean this is the only way to do things! We are confident that the information we’re providing is a good place to start; however, in practice, your style and approach to communication will evolve over time to best suit you, and that’s great! We recommend giving our approach a try at first. Then, as you gain experience, you can allow your own style to develop over time into whatever works best for you.

Additionally, we’ll do our best to provide links in each lesson to others’ thoughts on the topics we’re covering. This way, if you have time (either now or in the future), you’ll be able to see how others approach data science communication!

73.0.2 Types of Communication

Effective data science communication should always communicate your message clearly to your intended audience. This can be accomplished through a number of different types of communication. In this lesson we’ll be discussing the following general types of data science communication:

Reports
Presentations
Blog Posts
Meetings

73.0.3 Reports

Data science reports are a written form of communication used to summarize a data science project through text and figures. Written reports should always tell a story. What we mean here is that a finished report should have a beginning, middle, and an end. It should start with the question you’re answering and relevant background information (the beginning), continue with what you did for the project, which will often include both plots and text (the middle), and finish with a summary of your results (the end). Reports are most often sent to your teammates or manager to communicate what you’ve been working on.

73.0.4 Presentations

Presentations also tell a story; however, they do so in a different format and for a different audience. Presentations are generally comprised of slides with images and text on them that you present orally; however, they can include brief videos or GIFs as well. The audience for a presentation is generally larger than those who read your report. Presentations can be made to groups of individuals at your company or to a group of attendees at a conference.

73.0.5 Blog Posts

Blog posts are an effective way to reach a wide audience. Like reports, these also tell a story through text and figures; however, rather than update co-workers on your progress for a project at work blogposts are often on topics that are more helpful to a general audience. Data science blog posts are generally either “how to” posts that teach others how to do something related to data science or are “analysis” posts that summarize a cool analysis. Writing blog posts on your website or personal blog can be a great way to share you work with others in the community you don’t typically work with directly.

73.0.6 Meetings

Data scientists often find them in a number of meetings. Meetings can be with team members where you discuss the plan for a project. Or, they can be with someone who’s less familiar with data analysis but wants your help. Further, data scientists can end up in larger meetings with groups of people in a company where they as the data scientist are the only person familiar with the data or analysis in the room and are expected to communicate results to everyone. As a participant in all of these cases, it’s the job of the data scientist to communicate effectively with others in the room. In lessons in this course we will discuss how to communicate effectively in meetings as well as discuss how to both run a meeting and how to participate in someone else’s meeting.